I am using fabric scans for a couch and because the scanned area is so small one can begin to see patterns. Now I am using random offset with the triplanar option, so the direction of the fabric is always the same, but I have some variation, depending on the angle. But the pattern is visible on broader areas. Could a seed option for random offset on different objects be implemented, please? if the scanned area would be broader, it wouldnt be such a big problem, but these patterns take away the believeability of the material.
you are probably better off using poplin texture mask and mixing 3 materials inside of that.
Is that a regular Max material or so? Never heard of it
I’m not sure that the seed of the triplanar is the problem., the seed changes form object to object, but remains the same in single object, so if your couch is big enough and is single object, the random offset can’t help you
if you are using vray next, there is a better solution for this problem - the “runtime extending”. This feature allows to rearrange the “bricks” of the sample in a random way in ordero to avoid the repeated patterns.
The runtime extending is not kind of hidden option, so there is no way to enable it by a magic script words, it requires some build-time calculation stages, so we have to get your material, do some magics, and return it back to you.
So, if you want to try this solution, tell me which material is it, and where to upload the modified copy with RTE enabled
I actually have the couch separated into different parts/cushions so thats why I thought that changing seed for the offset would help. I tried making different materials and a coordinates offset, but that didnt seem to be applied.
the scans name is ft0020_s
thanks for helping me!
a place to upload? I could share a folder on google drive with you, but wouldnt it be more convenient if you used your own FTP?
I just wanted to check with you how fast you can send me the edited scan. Kind of on a tight schedule here. Thank you.
it’s ready, send me an invitation to gdrive folder
What’s your email address, please?
sent! thank you a lot!
Yse, if you are not using triplanar, the uv setup must be rearranged, the best way is to use the “adjust UV tiling” function
But to be honest i even can’t recognize the material on these shots, perhaps the color is changed, but the pattern of the third image is strange, perhaps this is kind of closeup view of the material
Are you sure thet you render with VRay Next? if you render this material with 3.6, the sample size is only 1.5 cm and there vill be well pronounced tiling. The RTE technique makes the effective sample size infinite, but the tiles in the first image can’t be explained in this case.
Yes, I changed the color. The first image is the size I am aiming for. With these different settings I cant find a setting that comes close to it. The adjustet UV tiling goes up to 30, which takes away all the details… with the original material I had it at 0.65, so I wonder why that would change…
I am definitely using vnext. let me try the newest nightly though..
Can you prepare simple scene (the couch only) and send it to me? (max 2017 )
sent to your email.
thanks
I think i found it - RTE is not implemented in GPU
Oh ok, too bad .
i could try to switch to CPU, but this could mess up a lot of things… Sigh
Would it be possible to receive the material as maps, please?
I could rearrange them per object.
No problem, i will upload them in the same folder
But beside this, i will try to recreate the material bigger in size, this should remove the repeating patterns.
Thank you .
how do you recreate it though? Rescan it? Or some other magic?
some other magic 
The recreated mtl is already uploaded, can’t see tiles in render, but apparently is few degrees rotated compared with the original one, to get the same result use 50 instead 45 deg
The textures are also uploaded but i find them not satisfying, the material is relatively old and the technology that we used back then is not friendly to the decomposition.